摘要: 森林冠层叶绿素含量直接反映着森林的健康和胁迫情况。叶绿素含量的准确估测,更是研究森林生态系统循环模型的关键。文章以PROSPECT+SAIL模型为基础,从物理机理角度反演森林冠层叶绿素含量。首先利用PROSPECT和SAIL模型模拟叶片水平和冠层水平的光谱,并建立叶片水平叶绿素含量的查找表反演叶片叶绿素含量,然后结合森林结构参数Leaf Area Index (LAI)实现叶片尺度与冠层尺度叶绿素含量的转化,从Hyperion影像反演研究区域冠层水平叶绿素含量。结果表明,叶绿素含量的主要影响波段为400~900 nm;PROSPECT模型模拟的叶片光谱和SAIL模型模拟的冠层光谱均与实测光谱拟合效果较好,相对误差分别为7.06%,16.49%;LAI反演结果的均方根误差RMSE=0.5426;利用PROSPECT+SAIL模型可以较好地反演森林冠层叶绿素含量,反演精度为77.02%。
关键词:高光谱遥感;叶绿素含量;Hyperion
Abstract:The forest canopy chlorophyll content directly reflects the health and stress of forest. The accurate estimation of the forest canopy chlorophyll content is a significant foundation for researching forest ecosystem cycle models. In the present paper, the inversion of the forest canopy chlorophyll content was based on PROSPECT and SAIL models from the physical mechanism angle. First, leaf spectrum and canopy spectrum were simulated by PROSPECT and SAIL models respectively. And leaf chlorophyll content look-up-table was established for leaf chlorophyll content retrieval. Then leaf chlorophyll content was converted into canopy chlorophyll content by Leaf Area Index (LAI). Finally, canopy chlorophyll content was estimated from Hyperion image. The results indicated that the main effect bands of chlorophyll content were 400-900 nm, the simulation of leaf and canopy spectrum by PROSPECT and SAIL models fit better with the measured spectrum with 7.06% and 16.49% relative error respectively, the RMSE of LAI inversion was 0.542 6 and the forest canopy chlorophyll content was estimated better by PROSPECT and SAIL models with precision=77.02%.
[1] Gitelson A A, Merzlyak M N, Lichtenthaler H K. Journal of Plant Physiology, 1996, 148: 501. [2] Carter G A. International Journal of Remote Sensing, 1994, 15: 697. [3] Sampson P H, Zarco-Tejada P J, Mohammed G H, et al. Forest Science, 2003, 49(3): 381. [4] Datt B. Journal of Plant Physiology, 1999, 154: 30. [5] Datt B. Remote Sensing of Environment, 1998, 66: 111. [6] Penuelas J, Baret F B, Filella I. Photosynthetica, 1995, 31: 221. [7] Blackburn G A. Remote sensing of Environment, 1998, 66: 273. [8] Haboudane D, Miller J R, Tremblay N, et al. Remote Sensing of Environment, 2002, 81: 416. [9] Zhang Y Q, Chen J M, Miller J R, et al. Remote Sensing of Environment, 2008, 112(7): 3234. [10] Renzullo L J, Blanchield A L, Guillermin R, et al. International Journal of Remote Sensing, 2006, 27(4): 817. [11] Zarco-Tejada P J, Miller J R, Noland T L, et al. IEEE Transactions on Geosciences and Remote Sensing, 2001, 39(7): 1491. [12] Zarco-Tejada P J, Miller J R, Harron H, et al. Remote Sensing of Environment, 2004, 89: 189. [13] Chen J M, Cihlar J. Remote Sensing of Environment, 1996, 55: 153. [14] Pearlman J, Carman S, Segal C, et al. In Proceedings of IGARSS, 2001, 17: 3036. [15] Liang S L, Fang H L, Thorp L, et al. Transactions on Geoscience and Remote Sensing, 2003, 41(1): 1260. [16] Datt B, McViear T R, VanNiel T G, et al. IEEE Transactions on Geoscience and Remote Sensing, 2003, 41(6): 1246. [17] Gondenough D G, Andrew D, Niemann K O, et al. IEEE Transactions on Geoscience and Remote Sensing, 2003, 41(6): 1321. [18] NIU Zheng, WANG Chang-yao, et al(牛 铮,王长耀,等). Remote Sensing and Applications for Carbon Cycle(碳循环遥感基础与应用). Beijing:Science Press(北京:科学出版社), 2008. 66. [19] Verhoef W. Ph.D. thesis. National Aerospace Lab., Amsterdam, The Netherlands, 1998. [20] Jacquemoud S, Baret F. Remote Sensing of Environment, 1990, 34: 75.